When trade accuracy is low due to contradictory or noisy indicators
When simplifying strategies to 3–4 core signals (EMA, RSI, MACD, optional Bollinger Bands)
When you need comparable confidence scoring across modules

Identify active indicators and hard gates in:
- modules/momentum.py
- modules/mean_reversion.py
- modules/ai_predictor.py
Reduce decision inputs to a coherent set:
- Trend: EMA spread (fast vs slow)
- Momentum: MACD histogram
- Mean reversion: Bollinger Bands + RSI (only when trend is weak)
- Volatility: ATR (risk sizing / SL/TP)
- Deep Learning: GRU model in prediction/predict.py consumes raw OHLCV sequences
Ensure indicators are computed consistently using shared utilities:
- utils/indicators.py
Preserve module output contract:
- {"signal": "BUY|SELL|HOLD", "confidence": float, "metadata": dict}

Más de este repositorio

mismo repositorio

Implements confidence-weighted signal combination with dynamic weights and execution thresholds

Structures ERC-8004-style logs (identity, reputation, validation, intent) and produces deterministic hashes

Aligns risk gating (SL/TP, sizing, volatility, caps) with signal confidence to avoid blocking good trades

Ensures trades are executed only when confidence and risk checks pass; logs trades deterministically

Builds a dashboard to visualize trades, portfolio, and agent reasoning

Generates trading strategy modules (momentum, mean-reversion, yield)

name	indicator_analysis
description	Audits indicators, removes contradictions, and simplifies signals to a small coherent set

When trade accuracy is low due to contradictory or noisy indicators
When simplifying strategies to 3–4 core signals (EMA, RSI, MACD, optional Bollinger Bands)
When you need comparable confidence scoring across modules

Identify active indicators and hard gates in:
- modules/momentum.py
- modules/mean_reversion.py
- modules/ai_predictor.py
Reduce decision inputs to a coherent set:
- Trend: EMA spread (fast vs slow)
- Momentum: MACD histogram
- Mean reversion: Bollinger Bands + RSI (only when trend is weak)
- Volatility: ATR (risk sizing / SL/TP)
- Deep Learning: GRU model in prediction/predict.py consumes raw OHLCV sequences
Ensure indicators are computed consistently using shared utilities:
- utils/indicators.py
Preserve module output contract:
- {"signal": "BUY|SELL|HOLD", "confidence": float, "metadata": dict}